Simulation Device and Simulation Method

The present invention relates to a simulation device and a simulation method for simulating a human flow.

In recent years, a human flow simulation system capable of reproducing a flow of pedestrians on a computer for the purpose of evaluating countermeasures against problems caused by local concentration of a crowd, such as prior evaluation of an evacuation guidance plan at the time of events or disasters or prior evaluation of a congestion situation accompanying construction of new commercial facilities has attracted attention.

In relation to the human flow simulation system, for example, PTL 1 proposes “obtaining a ratio of the number of moving agents proceeding in one direction and the number of moving agents proceeding in the other direction at a selected edge for each direction, calculating a width size relative to a direction in which the moving agent proceeds from the ratio with respect to a width of the edge, calculating, with respect to each of the moving agents, a population density based on an area of a section which is obtained from a length of a forward region from a position of the moving agent and the width size calculated for the direction in which the agent proceeds, and the moving agent existing in the section for each direction, and calculating, for each moving agent, a movement speed of the moving agent based on a free walking speed of agent information, the calculated population density, and a predetermined parameter” for the purpose of accurately simulating movement by simulating an interaction caused by a difference in proceeding direction.

PTL 1: WO 2021/044481 A

According to PTL 1, the walking speed is determined based on the number of moving people and the number of people moving in an opposite direction in the edge, but there is no mention of determination of a movement direction for avoiding a contact in a crowded situation. Therefore, for example, in an environment where an obstacle or a narrow portion, which becomes a bottleneck, exists, a complete deadlock sometimes occurs when an opposing flow to the bottleneck occurs.

In view of the above, an object of the present invention is to provide a simulation device and a simulation method capable of coping with a bottleneck and obtaining an acceptable human flow simulation result.

In view of the above, “a simulation device that simulates a human flow includes: a storage unit; and a control unit for simulation, in which the storage unit stores layout information and a human model that holds at least destination information indicating a destination in the layout information, and current position information, and the control unit determines a movement direction of the human model so as to avoid a region in front of a region where opposing persons having different destinations are lined up in succession”.

Further, in the present invention, “a simulation method for simulating a human flow includes: a storage method; and a simulation control method, in which in the storage method, an arbitrary layout and a human model that holds at least destination information indicating a destination in the layout, and current position information are stored, and in the control method, a movement direction of the human model is determined so as to avoid a region in front of a region where opposing persons having different destinations are lined up in succession.”

It is possible to provide a simulation device and a simulation method capable of coping with a bottleneck and obtaining an acceptable crowd flow simulation result.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

1 FIG. 1 12 13 14 is a diagram illustrating a configuration example of a simulation device according to a first embodiment of the present invention. A simulation devicehas a configuration in which a storage unit DB, a computation unit, and an input/output deviceare connected to a bus.

1 2 3 Among them, the storage unit DB stores layout information Dof an environment for which simulation is to be performed, a human model Dthat performs movement behavior under the environment, and an opposing person cost map Dset in a certain forward range.

2 In the human model D, information regarding a human model for which simulation is to be performed is stored for each person. Such information exists as many as the number of people, and include information such as a current position, a destination, and a walking speed of each person.

12 124 122 123 13 131 132 A functional processing content of the computation unitcan be said to include respective processing units including an opposing person cost map creation unit, a human movement direction determination unit, and a human movement execution unit. The input/output deviceincludes an input unitthat performs appropriate setting of a simulation condition or the like, and an output unit (display unit)that outputs input setting data and a simulation result.

1 FIG. 2 FIG. 3 a FIGS. 3 b. Hereinafter, an operation of the simulation device ofwill be described in detail, and several items will be described as a premise. First, as for the environment for which simulation is to be performed, an appropriate place or facility can be selected according to each analysis purpose, and in the embodiment, the inside of a building as illustrated inis taken as an example. As a method of estimating a movement direction of a person on a simulation model, it is assumed that the estimation is performed using a cost method inand

2 FIG. 2 FIG. The building illustrated in plan view ofas an example of the environment for which simulation is to be performed includes a plurality of doorways A and B. The doorway A faces, for example, an external passage or the like, and the doorway B is at a position where an elevator for moving to each floor in the building is installed. Although illustrated in a simplified manner here, a staircase and the like are additionally included as the doorways, and a plurality of doorways may be included. Since a security gate G or the like is installed between the doorways A and B in the building of, a region in plan view is divided into a passable region and a non-passable region.

2 FIG. 13 In order to perform human flow simulation for, for example, it is necessary to set conditions necessary for modeling movement of a person in the building in advance. A setting procedure roughly includes three steps. First, building layout setting in which a layout, a floor dimension, and a height of each floor in the building are set is performed with reference to an architectural drawing or the like. Next, various in-building facilities such as elevators, escalators, security gates, and automatic doors are arranged, and building facility setting in which a specification of each facility is set is performed. Then, an entrance and a doorway of a room are set, and human movement setting in which the number of people moving from one location to another location is set is performed. These steps of setting processing are performed in advance while performing appropriate display using the input/output device.

1 2 FIG. When the building layout setting and the building facility setting are completed, the layout information Din which the layout, the floor dimension, and the height of each floor in the building are set is formed and stored as illustrated in.

1 Furthermore, in the human movement setting on the premise of the above setting, the entrance and the doorway of the room are set, and the number of people moving from one location to another location is set. Information regarding the entrance and the doorway of the room is included in the layout information Dand managed.

3 a FIG. 3 a FIG. 1 Here, as for handling of the human model at this time, in particular, the cost method is adopted in the present invention as a concept of determining a movement direction of the human model.is a diagram illustrating the concept of determining the movement direction of the human model in a normal state. Here, for example, eight directions (a front direction, a rear direction, a left direction, and a right direction, and a left-front direction, a right-front direction, a left-rear direction, and a right-rear direction) are set on a floor plane as the movement direction of the human model M, and a cost is set for each direction. The cost is set in terms of magnitude. Normally, in a case where 10 is set for the front direction, a numerical value larger than 10 is set for other directions. In, only the front direction, the left-front direction, and the right-front direction are simply described, and 14, which is larger than 10, is set for the left-front direction and the right-front direction. In this setting, a direction for which the cost is lowest is set as a proceeding direction.

2 3 b FIG. In addition, in actual simulation on the premise of such cost setting, when an opposing person Mis detected ahead in the movement direction as illustrated in, a cost for a detection direction is adjusted to be higher than an initial setting value of 10, for example, 15. As a result, it is possible to set the movement direction of the human model so as to perform the avoidance action in the left-front direction and the right-front direction or the like according to the principle of setting the direction for which the cost is the lowest as the proceeding direction.

The above is the premise of the simulation technique in the present invention. In the case of the above model, the human model in front of the opposing person can perform the avoidance behavior, but when many human models gather in a narrow portion such as the bottleneck, the occurrence of the avoidance behavior is slow and the avoidance cannot be made, as a result of which a complete deadlock in which a group cannot move due to collision with an opposing group may occur.

Hereinafter, processing in the simulation device in the present invention will be described.

124 12 3 After preparing the contents to be stored, the opposing person cost map creation unitin the computation unitcreates the opposing person cost map Dfor each destination of the human model.

4 4 a b FIGS.and 4 a FIG. 4 b FIG. 3 1 21 22 3 2 1 21 22 3 are diagrams illustrating a state in which the opposing person cost map Dcorresponding to a person heading for an exit is created. In these figures, the person heading for the exit is represented as M, and persons heading for an entrance are represented as Mand M. The opposing person cost map Dis configured by housing a cost value CMof the opposing person cost map for each point with respect to a certain forward range CMother than the person heading for the exit. In, a cost value of 3 is added to three points in front of the human model M. In, a cost value of 3 is added to three points ahead of the human model M. In this manner, the cost value is added to a region in front of a person heading for other than a target destination, whereby the opposing person cost map Dof the target destination is created.

3 The opposing person cost map Dmay be created not for each destination of the human model but for each movement direction candidate. In a case where there are many destinations, it is expected that a calculation cost and a storage cost can be reduced by this method.

When setting the cost value for each point in the region in front of the person, the cost value may be set to be increased at a point closer to the opposing person.

Furthermore, the cost value added to each point in the region in front of the person may be set to a value smaller than a difference between the smallest value and the second smallest value of a distance cost value for each movement direction of the person. In the case of setting in this manner, the distance cost for another direction becomes larger than that for a direction for which the distance cost becomes the smallest only in a case where a plurality of opposing persons are lined up in succession. Therefore, for example, in a case where one person and a group forming a line face each other, the one person changes the proceeding direction to another direction instead of the group changing the proceeding direction for the one person, so that it is possible to reproduce a more acceptable motion.

2 Here, for the cost value CMof the opposing person cost map, in a case where the opposing person is in the vicinity of a non-passable region such as the bottleneck, processing such as increasing an additional cost may be performed. As a result, an effect of making it easy to cause the avoidance behavior in a narrow region is expected.

124 12 3 2 1 FIG. As described above, the opposing person cost map creation unitin the computation unitinperforms opposing person cost map creation processing for each destination in each operation cycle of the simulation, and reflects the creation result in the opposing person cost map D. The reflected content is the cost value CMof the opposing person cost map for each point and the like.

122 5 5 5 a b c FIGS.,, and 5 a FIG. 3 a FIG. Next, processing in the movement direction determination unitwill be described with reference to. The movement direction determination unit determines the next movement direction of each human model in each operation cycle of the simulation.illustrates an example of a magnitude of the cost for each direction in the same normal state as, and the cost for the other directions is set to 14 higher than 10 which is the cost for the front direction. Therefore, according to the principle of selecting the lower one, the front direction is set as the proceeding direction in the normal state.

5 b FIG. 2 illustrates an example in which there is only one opposing person in the front direction. +3 is added to the cost for the front direction in the normal state with reference to the cost value CMof the opposing person cost map, and thus, a movement cost for the front direction becomes 13. However, the movement cost is lower than the cost of 14 for the peripheral directions, the front direction is still selected as the proceeding direction in this case.

5 c FIG. 2 1 1 illustrates an example of a case where there is still another opposing person. The cost value CMof the opposing person cost map in front of the human model Mis 6 and is added to the movement cost for the front direction in the normal state, and thus, the movement cost for the front direction becomes 16. At this time, since the movement cost for the front direction becomes 16 higher than 14 which is the cost for the peripheral directions for the first time, one of the left-front direction and the right-front direction is selected as the proceeding direction, and direction change is performed. In a case where the cost value is the same between the movement directions, selection of one of the left direction and the right direction may be performed using a random number or may be performed based on cultural area information input in advance to the layout D. For example, the right-front direction may be preferentially selected in Japan which follows a right-side walking rule.

122 The movement direction determination unithas been described above. According to the above processing, the person selects the movement direction so as to avoid a region in front of a point where the opposing persons are lined up in succession, so that movement in which a movement space of the opposing persons is secured can be implemented. Therefore, a complete deadlock can be avoided. This leads to an improvement in accuracy of a simulator and satisfaction for the motion.

123 123 122 Finally, the human movement execution unitwill be described. The human movement execution unitchanges a position of the human model based on a movement direction of a person determined by the human movement direction determination unit. In this way, human flow simulation is implemented by repeating the determination of the movement direction and the position change of the human model for each simulation operation cycle.

In the description of the premise of the above cost setting, the cost for the front direction is set to be lower than the cost for other directions, a cost increase operation is performed according to the presence of the opposing person or the bottleneck, and a direction for which the cost is low is set as the proceeding direction. However, a magnitude relationship thereof and the like may be reversed. That is, the same applies even in a case where the cost for the front direction is set to be higher than the cost for other directions, a cost decrease operation is performed according to the presence of the opposing person or the bottleneck, and a direction for which the cost is high is set as the proceeding direction.

3 3 As described above, in the present invention, the opposing person cost map Dis prepared for each destination of a person, and the cost value is added to each point in the region of in front of the opposing person. In the direction determination, the movement direction is determined with reference to not only a distance to the destination but also a value of the opposing person cost map D. As a result, since the motion can be changed only when a congestion level is increased, it is possible to make a motion preventing a deadlock at a narrow portion such as the bottleneck without changing human flow characteristics at other points. This leads to an improvement in accuracy of a simulator and satisfaction for the motion.

6 FIG. 7 FIG. In a second embodiment, a simulation method will be described.illustrates a processing flow in a prior preparation stage, andillustrates a processing flow during simulation.

11 12 13 1 14 6 FIG. Building layout setting is performed in the first processing step Sin, building facility setting is performed in processing step S, and human movement setting is performed in processing step S, that is, the number of people moving for each departure point and destination point is set for each time. By execution of a series of these steps of processing, each piece of layout information Dis formed and stored in processing step S.

7 FIG. 21 1 2 3 22 2 29 In, which illustrates the processing during the simulation, first, in processing step S, the layout information Dand a human model Dof an environment in which the human flow processing is performed, opposing person cost map creation data D, and the like are obtained. In processing step S, repetitive processing is performed while changing a human model in the human model Duntil processing step S. It is assumed that the environment and thus a layout are specified.

23 25 24 2 In the repetitive processing, in processing step S, the repetitive processing is performed while changing a movement direction candidate until processing step S. In processing step S, a movement cost value Cmove corresponding to the specified movement direction candidate calculated by obtaining the sum of a distance cost C and a cost value CMof an opposing person cost map corresponding to a destination of a target human model M.

25 24 Processing step Sis termination determination processing for iteration, and processing step Sis repeated a number of times corresponding to the number of movement direction candidates.

26 In processing step S, the movement cost value Cmove for each movement direction candidate is compared, and a movement direction for which the movement cost value is the smallest is selected to determine the movement direction of the human model M.

27 Processing step Sis termination determination processing for iteration, and the above processing is repeated for the human models corresponding to the set number of people.

The processing flow during the simulation has been described above. By repeating the above processing for each simulation operation cycle, the movement direction of the person is determined again so as to avoid the opposing person every time the position of the human model is changed as time passes.

1 simulation device DB storage unit 12 computation unit 13 input/output device 14 bus 1 Dlayout information 2 Dhuman model 3 Dopposing person cost map 1 CMcertain forward range 2 CMcost value of opposing person cost map 122 human movement direction determination unit 123 human movement execution unit 124 opposing person cost map creation unit 131 input unit 132 output unit (display unit)